Capacity Planning for Growing Manufacturers

Capacity planning is how you work out whether you have the machine hours and labour hours to deliver the work you’ve already accepted — and whether you can safely take on more. It compares the load coming at you (every job, its hours, its due date) against the capacity you actually have (people, machines, shifts, minus the time lost to setup, maintenance and absence). When those two numbers are close, you’re fine. When load quietly climbs above capacity, jobs go late — and the first sign is usually a customer phoning, not a report.

This isn’t the same as scheduling. Scheduling decides the order jobs run once you’ve committed to them; that’s a separate job covered in manufacturing scheduling software. Capacity planning sits upstream of that: it answers “do we even have the hours for this workload?” before you promise anything. Get it wrong and no amount of clever sequencing saves you — you’ve sold hours you don’t own. This post walks through load versus capacity, finite versus infinite planning, spotting the bottleneck, rough-cut capacity planning, and using the whole thing to quote a date you can keep.

Key Takeaways

Capacity planning compares load (the hours of work you’ve accepted) against available capacity (real machine and labour hours after setup, maintenance and absence) — not nameplate hours.
Most missed delivery dates are a capacity failure that was invisible at the moment someone said “yes,” not a scheduling mistake later.
Available capacity is always less than theoretical capacity. Plan on the hours you truly get, or every plan runs optimistic.
Finite capacity planning refuses to load a work centre past its limit; infinite capacity planning lets it overflow and shows you the overload. Both are useful — for different questions.
Your bottleneck sets your real output. Adding work anywhere else just builds queue in front of it.
Rough-cut capacity planning — a fast check against your one or two constraint resources — is enough to accept or decline most orders and quote an honest date.

1What Capacity Planning Actually Answers

Every quote you send is a promise about hours you may not have counted. Capacity planning is the counting. On one side you put load: every job on the books, the hours it needs at each work centre, and when it’s due. On the other you put capacity: how many productive hours those work centres can actually deliver in the same period. Line them up week by week and the picture is blunt — either the work fits in the time, or it doesn’t.

That’s the whole discipline, and most growing manufacturers do it in their head. It works until it doesn’t. One production planner we spoke to described the moment it broke: “We said yes to three good orders in the same fortnight because each one, on its own, looked fine. Nobody added them up against the hours we actually had. All three ran late — and they were all late for the same week.” No single decision was wrong. The sum was never checked. That’s the gap capacity planning fills: it makes the total visible before you commit, not after the customer calls.

2Load vs Available Capacity — Count the Hours You Truly Get

The number that wrecks capacity plans is the one people assume. A machine “runs 40 hours a week,” so that’s the capacity — except it isn’t. Take off setup and changeover, planned maintenance, the operator’s breaks and holidays, the jobs that need rework, the mornings the material’s late. What’s left — the hours the machine actually turns out good parts — is your available capacity, and it’s routinely well short of the nameplate figure, not the full 100%. Plan on 40 when you get 28 and every week runs a third short before you’ve done anything wrong.

The fix is unglamorous: measure what you really get. Track how many productive hours each work centre delivered last month, not what the spec sheet claims. Feed that real number into the plan. A production monitoring system does this for you by logging actual run time against each machine, so your available-capacity figure comes from the floor instead of an optimistic guess. Load then goes on the other side of the ledger: sum the standard hours of every accepted job by the week it’s due. Two honest numbers, compared. That comparison is the entire game.

3Finite vs Infinite Capacity Planning

There are two ways to load work against capacity, and knowing which you’re doing keeps you honest. Infinite capacity planning assumes every work centre can take whatever you throw at it — it schedules jobs to their due dates and lets a resource go to 130% loaded without complaint. That sounds useless, but it’s exactly what you want for seeing the problem: the overload shows up as a bar poking above the capacity line, screaming “this week is impossible.” It diagnoses. It doesn’t fix.

Finite capacity planning refuses to overload. It caps each work centre at its real limit and pushes the overflow work into the next available slot — so the plan it produces is always achievable, but some jobs land later than their original due date. That’s the trade: infinite tells you the truth about demand, finite tells you the truth about delivery. Growing shops usually want both — infinite to spot the overload early, finite to see the honest completion dates once reality bites. The mistake is running infinite in your head, seeing due dates you’ll never hit, and promising them anyway.

4Find the Bottleneck — It Sets Your Real Output

A factory manager described the lesson the hard way: “We bought a second CNC to speed things up. Output didn’t move. The jobs just piled up faster in front of the paint booth — that was the real limit, and we’d spent forty grand nowhere near it.” Your output isn’t set by your average capacity. It’s set by your single tightest resource — the bottleneck. Every hour you add elsewhere just grows the queue in front of the constraint. Capacity planning that doesn’t identify the bottleneck is measuring the wrong thing.

Finding it is simpler than it sounds: the bottleneck is the work centre with the highest load-to-capacity ratio — the one running closest to, or over, 100% while others have slack. Look at your load-versus-capacity picture across all work centres and the constraint is the tallest bar. Once you know it, capacity planning gets sharper, because the bottleneck’s capacity is your factory’s capacity. Protect its hours, keep it fed, don’t waste its time on setups you could batch — and never accept an order without checking what it does to that one resource. Everything else has slack to absorb a bad week. The bottleneck doesn’t.

5Rough-Cut Capacity Planning — Fast Enough to Quote With

You don’t need a full plant model to answer “can we take this?” Rough-cut capacity planning is the quick version: instead of loading every operation across every machine, you check the new order against your one or two constraint resources only. If a job needs 30 hours on the bottleneck and the bottleneck has 22 free hours before the customer’s date, you have your answer without modelling anything else. It’s approximate on purpose — approximate and fast beats precise and too late.

This is the level most growing manufacturers actually need. It’s enough to accept or decline an order at the moment you’re quoting, and enough to put an honest date on the quote instead of a hopeful one. Rough-cut sits above the detailed materials logic of an MRP system — you’re not exploding bills of materials, you’re sanity-checking the hours against your tightest constraint. Do it on the constraint resources and you catch the overloads that cause almost all your late deliveries, in the time it takes to send the quote.

6From Capacity Check to Honest Delivery Date

Here’s where capacity planning earns its keep in pounds. An overpromised order costs you twice: once in the rush freight, overtime and rework to claw it back, and again in the customer who won’t quote you next time because you were three weeks late. A shop that quotes off a real capacity check quotes dates it hits — and a manufacturer known for hitting dates can charge for it. The honest longer date wins more repeat business than the optimistic one you break.

The before-and-after is concrete. Before: sales promises a four-week lead time because that’s what it usually is, nobody checks the bottleneck is already loaded to week six, the order lands late, and everyone firefights. After: the same order gets checked against constraint capacity at quote time, comes back “six weeks, honestly,” and either the customer accepts a date you keep or you flag the overload and decide whether to add a shift. Either way you chose with the numbers in front of you. That decision — take it, decline it, or add capacity for it — is the entire point of capacity planning. The scheduling of how those committed jobs then run is the next layer, handled by production tracking and scheduling once the promise is real.

7Why Spreadsheets Break at This Job

Capacity planning is exactly the kind of maths a spreadsheet flatters you into thinking you’ve solved. You can build a grid of work centres and weeks, drop in loads, colour the overloads red — and it works, for one snapshot, maintained by one person. The problem is that load changes every time an order lands, a job slips, or a machine goes down, and the sheet only updates when someone remembers to update it. By the time the red cell appears, you’ve already promised the order that caused it.

The other break is the input. A capacity plan is only as true as its available-capacity figure, and a spreadsheet can’t know the CNC actually ran 26 hours last week, not 40 — someone has to type it, and nobody has time. So the sheet drifts to nameplate hours and quietly goes optimistic. A built-for-you system closes both gaps: it pulls real run hours from the floor, updates load live as orders and job status change, and shows load against real capacity per work centre without anyone maintaining a grid. You get the constraint flagged before you quote, not after you’re late. That’s the difference between capacity planning as a document and capacity planning as a live decision tool.

FAQ

What is capacity planning in manufacturing?

It’s the process of comparing the work you’ve accepted (load — the machine and labour hours every job needs) against the hours your work centres can actually deliver (available capacity). Done week by week, it tells you whether the workload fits in the time you have. Its main job is to let you decide whether you can accept an order and quote a delivery date you’ll actually hit, before you commit to the customer.

What’s the difference between capacity planning and scheduling?

Capacity planning is upstream: it asks whether you have enough total hours for the workload, and whether to take a new order at all. Scheduling is downstream: once you’ve committed to the jobs, it decides the order and timing they run in on each machine. You can have a perfect schedule and still miss every date if you sold more hours than you own — that’s a capacity failure, not a scheduling one.

What is the difference between finite and infinite capacity planning?

Infinite capacity planning loads jobs to their due dates and lets a work centre go over 100% — useful for seeing an overload, because the excess shows up clearly. Finite capacity planning caps each resource at its real limit and pushes overflow into the next free slot, so the plan is always achievable but some completion dates move later. Infinite tells you the truth about demand; finite tells you the truth about deliverable dates.

How do I find my bottleneck?

It’s the work centre with the highest load-to-capacity ratio — the one running closest to or above 100% while others have slack. On a load-versus-capacity view across all your work centres, it’s the tallest bar. That resource sets your real output, so adding capacity anywhere else just builds queue in front of it. Check every new order against the bottleneck first.

Do I need software for capacity planning, or is a spreadsheet enough?

A spreadsheet works for a single snapshot maintained by one person, but load changes with every new order and every slipped job, and the sheet only updates when someone remembers. It also can’t see real run hours, so it drifts to optimistic nameplate figures. A built-for-you system pulls actual capacity from the floor and updates load live, so the overload is flagged before you quote — not after you’re already late.

How OpsMavix Can Help

OpsMavix builds right-sized capacity planning into the production tracking systems we make for growing manufacturers — the shops winning orders faster than they can honestly deliver them, or watching one machine drown while another sits idle. Instead of a spreadsheet that’s true for a day and optimistic by the next order, you get a live view of load against your real available capacity per work centre: your constraint flagged, your bottleneck obvious, and a quick rough-cut check you can run at quote time to accept, decline or price an order with the hours in front of you. It pulls actual run time from the floor so the capacity figure is real, updates as jobs land and status changes, and connects to the tools you already run. You own it outright, no per-seat fees, live in weeks.

If you’re quoting dates you can’t keep, or you can’t see where your load and your real capacity part ways, start by finding the leak. Book a Free Operations Leak Audit and we’ll map where your accepted work outruns your actual hours, what the overpromising is costing you in rush freight and lost repeat business, and whether a right-sized capacity view fixes it.